Getting Started with Git

Leyang Feng
June 12, 2019

A workshop covering reproducibility; version control; basic git workflows.

Reproducibility

Reproducibility

We are in the era of 'big data', but even if you work with 'little data' you have to acquire some skills to deal with those data.

Most fundamentally, your results have to be reproducible.

Your most important collaborator is your future self. It’s important to make a workflow that you can use time and time again, and even pass on to others in such a way that you don’t have to be there to walk them through it. Source

Reproducibility means scripts or programs tied to open source software.

You can't reproduce

…what doesn't exist.

  • Gozilla ate my computer
  • backup
  • ideally continuous
  • Godzilla ate my office
  • cloud

You can't reproduce

…what you've lost. What if you need access to a file as it existed 1, 10, or 100, or 1000 days ago?

  • Incremental backups (minimum)
  • Version control (better). A repository holds files and tracks changes

Version control

What on earth did I do???

Git and GitHub

Git (and website GitHub) are the most popular version control tools for use with R, and many other languages:

  • version control
  • sharing code with collaborators
  • issue tracking
  • social coding

Git

  • a version control software (stand along)
  • standard installation for most Unix/Linux distributions
  • standard installation for Mac OSX with Xcode developer tools
  • create snaphot for changed file in your folder/repository.

Git workflow

  • set up your working directory (folder)
  • initialize git in your folder

  • file operations in your folder (create/change/delete)

    just remember **git is watching you**

  • commit then changes so that the git can create current snapshot of your folder

Git to GitHub

Using Git, now you have version control, but you still have:

  • Godzilla
  • hard disk failure
  • …

Or, your want to:

  • collaborate with others that you cannot just transfer the code with a thumb drive
  • simply show-off your cool/fancy/geniusly written code

GitHub

  • a platform that provides remote repository storage
  • integration with git to track code change
  • built-in tools for collaboration
  • monitor stats of your repository

Basic GitHub workflow

  • start/clone a repository from GitHub (creating folder and initialize git)
  • file operations in your folder (create/change/delete)
  • commit then changes so that the git can create current snapshot of your folder
  • push the changes to GitHub repository

Team Work (or not) on GitHub -- Branching

pull requests…

Pull Request

  • propose to merge branch to master
  • forum style discussion board

Putting all together

Reproducible research example

A typical project/paper directory for me:

1-download.R
2-process_data.R
3-analyze_data.R
4-make_graphs.R
logs/
output/
rawdata/

This directory is backed up both locally and remotely, and is under version control, so it's easy to track changes over time.

Reproducible research example

  • Sequentially numbered R scripts (0-download.R, 1-process_data.R, …)
  • Each script depends on the previous one
  • Each has a discrete, logical task
  • Each produces a log file including date/time, what R version was used, etc.
  • This analytical chain starts with raw data
  • …and ends with figures and tables for ms
  • or the ms itself! This presentation, for example, was generated directly from an R file

Using git

  • GUI client
  • CLI

GUI client

  • GitHub desktop

  • SourceTree

CLI

Other Resources

Getting Started with Git

GitHub Desktop tutorials

Git Pro